[SOLVED] Unusual cpu fan wiring in laptop

[Mushmo]

Update:
After getting the tool I needed I was able to remove the cover from my laptop fans and see their circuitboards. This let me see for certain that while the two GPU fans use three wires (power, ground and sense), the cpu fan is using power, ground and pwm, with the pwm wire plugged into the sense slot of its header, and the sense wire unconnected to anything.


Original post:
So I noticed that of the three fans in my Metabox P950HP laptop(specs), one seemed to be moving almost no air compared to the other 2, even at extremely high temps (high 80s, Celsius). The laptop has one fan for the cpu and 2 fans for the gpu. Both of the gpu fans are 3 wire fans, and the cpu fan is a 4 wire fan but it goes through a 4 pin to 3 pin adapter to fit into the motherboards 3 pin cpu fan header.

The two 3-wire gpu fans have red, black and yellow wires. If I understand right, black is always ground and red is almost always power, so yellow must be sense/tach. The cpu fan has the same, plus an additional white wire, which I think must be the PWM/control wire.

Here's the thing that confuses me: after converting from 4 wires to 3 the cpu fan uses the black red and white wires, while the other 2 fans use red black and yellow. In other words, I think that the cpu fan is using ground, power and control/pwm, but I thought that all 3-wire fans use ground, power and sense/tach?

I also noticed that in the 4 to 3 adapter, on the 3 side, where the control wire exits the adapter seems to be burned maybe? And there seems to be a very small hole or something in the centre of the burn.

All of the fans are the same brand, with the same ordering of the wires in terms of colour. I found the laptop schematic and the third slot in all three fan headers (after power and ground) is supposed to be sense/tach.

I contacted the manufacturer (Metabox), and they confirmed that this wiring for the cpu fan is definitely correct.

Can anyone explain this? Is it possible for a fan to use control and not sense/tach?

I made a diagram showing the wiring setup to better explain the wiring layout.

To summarize, one of my 3 laptop fans has the pwm/control wire plugged into the sense/tach plug in the 3 pin header, and leaves the actual sense/tach wire unconnected. The other two fans are 3-wired and so use the sense/tach wire, as is typical for 3 wires fans.

I've been trying to figure this out for a week, any advice would be much appreciated, thanks

Fans
Laptop Manual (fans on page 102)
Pictures

 

[lvt]

If your laptop's motherboard doesn't have PWM then the fans don't need 4-pin connectors.

The fans's speed can also be controlled by voltage regulation.

For a laptop, using voltage regulation might be more battery efficient as PWM uses constant voltage regardless of fans speed.

 

[Mushmo]

Thanks for your response

Yeah I understand that 4, 3 and even 2 pins are viable and how each works (to a degree). I'm not concerned that the cpu fan is using 3 wires, rather that it's using the wrong 3 wires. The PWM wire, which I believe sends input to the fan, is plugged into the sense port in the header, which recieves output from the fan. I don't understand how that works, or if it even does work? Also do you have any thoughts on the damage to the pwm wire?

 

[lvt]

Thanks for your response

Yeah I understand that 4, 3 and even 2 pins are viable and how each works (to a degree). I'm not concerned that the cpu fan is using 3 wires, rather that it's using the wrong 3 wires. The PWM wire, which I believe sends input to the fan, is plugged into the sense port in the header, which recieves output from the fan. I don't understand how that works, or if it even does work? Also do you have any thoughts on the damage to the pwm wire?

The problem is that in your assumption, the yellow is Tach and the white is PWM, but we can't confirm that without actually seeing the fan's circuit board. The wires can be wired differently according to buyer's request as well. For example, in this case, to match the laptop's specs, both yellow and white wires might be Tach wires (soldered to the same spot).

And there are fake cooling fans that run solely on the red and black wires, the other 2 wires do absolutely nothing.

 

[Paperdoc]

I agree with IVT. You can NOT presume that wire colour codes are being used in the most common way with every fan. Nor can you presume that a 4-pin fan means it is a PWM design. And you are right - irrespective of the fan colours and wire counts, the manual says ALL fan headers are 3-pin, and that means they can ONLY provide power and control of the fan by the older Voltage Control Mode. It is very likely that even the 4-wire fan is done that way, and the white wire is actually the fan's speed output signal. In fact, here's an idea to check. In BIOS Setup does it actually show you a speed reading for that CPU cooler fan? If yes, then some wire IS carrying that signal to the mobo.

If you see NO CPU fan speed, the speed signal is not getting through. You may have told us an important clue. You say that, at the output of the connector where wire connections change, there appears to be a flaw in the wiring for the Yellow wire. At very minimum (with power off, of course) you should disconnect that wire segment (from that connector to the mobo header) and check resistance from connector contacts at one end to the other. If there is any reason to suspect poor continuity on the Yellow wire, repair the connection. Also look for some reason why there MAY be a hole in the insulation, and fix that. After re-assembly, verify that a CPU fan speed is shown.

By the way, FYI, the design of PWM fans includes a backwards compatibility feature. If it is connected to a header using the older 3-pin fan Voltage Control Mode, it receives NO PWM signal from Pin #4 (that does not exist in the header) so its internal chip cannot use that to modify power flow from Pin #2 (the +VDC power source) through the windings to adjust fan speed. However, unlike a normal PWM mode system, the older system supplies on Pin #2 a DC Voltage that VARIES from 12 V (full speed) to about 5 V (minimum to avoid fan stalling). this DOES accomplish speed control of the fan even though it is not ideal because it fails to use the PWM design. So, whether or not that CPU fan is a 4-pin PWM design, it CAN function properly from the header signals it gets, provided the three wires in use are connected correctly.

 

[Mushmo]

Hi, thanks for taking the time to respond!

You can NOT presume that wire colour codes are being used in the most common way with every fan. Nor can you presume that a 4-pin fan means it is a PWM design

Yeah you're totally right, different brands use different colours. I've also seen some cheap fans that have four wires but the extras don't connect to anything . Even wire order on the fan isn't standard, I think that Dell is infamous for using their own standard that differs from most others.

Yellow being tach seemed the most reasonable assumption for three reasons:

1. The 2 GPU fans both plug their yellow wires into the tach port of their respective headers.
2. The GPU fans only have 3 wires, and assuming the red and black wires are power and ground which is almost always the case, yellow must be tach.
3. The yellow wire exits the same part of the CPU fan as it does the GPU fans, leading me to believe it performs the same function.

In BIOS Setup does it actually show you a speed reading for that CPU cooler fan? If yes, then some wire IS carrying that signal to the mobo

The only program that gives me any kind of fan speed reading is SpeedFan, which shows one reading labelled "CPU Fan" that ranges from roughly 40,000 to around 6,500,000 RPM or so... I don't know how to explain such a high number but I know it has to be inaccurate... The Nvidia overlay shows a fan speed of zero, and open hardware monitor shows no fans. In BIOS I can't find anywhere to view fan speeds, this laptop’s BIOS interface is extremely basic.


I just wanted to clarify all of that and add extra info before giving an update: I just got my hands on a screwdriver small enough to remove the extra small screws holding the fan covers in place, and I can now see the fan’s circuit boards. It looks like my assumption was right; on all 3 fans yellow is tach (it's labelled "FG" on the circuit board, I think that stands for feedback generator), and the fourth connection is labelled PWM. This means that for certain the cpu fan is connecting it's PWM wire to the tach port in the header.

Is there any situation where that would be something you'd want? And could the wire damage be caused by an input and output wire being connected to each other?

 

[Mushmo]

The problem is that in your assumption, the yellow is Tach and the white is PWM, but we can't confirm that without actually seeing the fan's circuit board. The wires can be wired differently according to buyer's request as well. For example, in this case, to match the laptop's specs, both yellow and white wires might be Tach wires (soldered to the same spot).

And there are fake cooling fans that run solely on the red and black wires, the other 2 wires do absolutely nothing.

Hey thanks for getting back to me, if you’re interested see the response and update I left to Paperdoc below